1
Stanley Fish: “Is There a Text in This Class?”
On the first day of the new semester, a colleague at Johns Hopkins University was approached by a student who, as it turned out, had just taken a course from me. She put to him what I think you would agree is a perfectly straightforward question: "Is there a text in this class?" Responding with a confidence so perfect that he was unaware of it (although in telling the story, he refers to this moment as "walking into the trap"), my colleague said, "Yes; it's the Norton Anthology of Literature," whereupon the trap (set not by the student but by the infinite capacity of language for being appropriated) was sprung: "No, no," she said, "I mean in this class do we believe in poems and things, or is it just us?" Now it is possible (and for many tempting) to read this anecdote as an illustration of the dangers that follow upon listening to people like me who preach the instability of the text and the unavailability of determinate meanings; but in what follows I will try to read it as an illustration of how baseless the fear of these dangers finally is.
Of the charges levied against what Meyer Abrams has recently called the New Readers (Jacques Derrida, Harold Bloom, Stanley Fish) the most persistent is that these apostles of indeterminacy and undecidability ignore, even as they rely upon, the "norms and possibilities" embedded in language, the "linguistic meanings" words undeniably have, and thereby invite us to abandon "our ordinary realm of experience in speaking, hearing, reading and understanding," for a world in which "no text can mean anything in particular" and where "we can never say just what anyone means by anything he writes." The charge is that literal or normative meanings are overriden by the actions of willful interpreters. Suppose we examine this indictment in the context of the present example. What, exactly, is the normative or literal or linguistic meaning of "Is there a text in this class?"
Within the framework of contemporary critical debate (as it is reflected in the pages, say, of Critical Inquiry) there would seem to be only two ways of answering this question: either there is a literal meaning of the utterance and we should be able to say what it is, or there are as many meanings as there are readers and no one of them is literal. But the answer suggested by my little story is that the utterance has two literal meanings: within the circumstances assumed by my colleague (I don't mean that he took the step of assuming them, but that he was already stepping within them) the utterance is obviously a question about whether or not here is a required textbook in this particular course; but within the circumstances to which he was alerted by his student's corrective response, the utterance is just as obviously a question about the instructor's position (within the range of positions available in contemporary literary theory) on the status of the text. Notice that we do not have here a case of indeterminacy or undecidability but a determinacy and decidability that do not always have the same shape and that can, and in this instance do, change. My colleague was not hesitating between two (or more) possible meanings of the utterance; rather, he immediately apprehended what seemed to be an inescapable meaning, given his prestructured understanding of the situation, and then he immediately apprehended another inescapable meaning when that understanding was altered. Neither meaning was imposed (a favorite word in the anti-new- reader polemics) on a more normal one by a private, idiosyncratic interpretive act; both interpretations were a function of precisely the public and constituting norms (of language and understanding) invoked by Abrams. It is just that these norms are not embedded in the language (where they may be read out by anyone with sufficiently clear, that is, unbiased, eyes) but inhere in all institutional structure within which one hears utterances as already organized with reference to certain assumed purposes and goals. Because both my colleague and his student are situated in that institution, their interpretive activities are not free, but what constrains them are the understood practices and assumptions of the institution and not the rules and fixed meanings of a language system.
Another way to put this would be to say that neither reading of the question-which we might for convenience sake label as "Is there a text in this class?"1 and "Is there a text in this class?"2-would be immediately available to any native speaker of the language. "Is there a text in this class?"1 is interpretable or readable only by someone who already knows what is included under the general rubric "first day of class" what concerns animate students, what bureaucratic matters must be attended to before instruction begins) and who therefore hears the utterances under the aegis of that knowledge, which is not applied after the fact but is responsible for the shape the fact immediately has. To someone whose consciousness is not already informed by that knowledge, "is there a text in this class?"1, would be just as unavailable as "is there a text in this class?"2 would be to someone who was not already aware of he disputed issues in contemporary literary theory. I am not saying that for some readers or hearers the question would be wholly unintelligible (indeed, in the course of this essay I will be arguing that unintelligibility, in the strict or pure sense, is an impossibility), but that there are readers and hearers for whom the intelligibility of the question would have neither of the shapes it had, in a temporal succession, for my colleague. It is possible, for example, to imagine someone who would hear or intend the question as an inquiry about the location of an object, that is, "I think I left my text in this class; have you seen it?" We would then have an "Is there a text in this class?"3 and the possibility, feared by the defenders of the normative and determinate, of an endless succession in numbers, that is, of a world in which every utterance has an infinite plurality of meanings. But that is not what the example, however it might be extended, suggests at all. In any of the situations I have imagined (and in any that I might be able to imagine) the meaning of the utterance would be severely constrained, not after it was heard but in the ways in which it could, in the first place, be heard. An infinite plurality of meaning would be a fear only if sentences existed in a state in which they were not already embedded in, and had come into view as a function of some situation or other. That state, if it could be located, would be the normative one, and it would be disturbing indeed if the norm were free-floating and indeterminate . But there is no such state; sentences emerge only in situations, and within those situations, the normative meaning of an utterance will always be obvious or at least accessible, although within another situation that same utterance, no longer the same, will have another normative meaning that will be no less obvious and accessible. (My colleague's experience is precisely an illustration). This does not mean that there is no way to discriminate between the meanings an utterance ca have in different situations, but that the discrimination will already have been made by virtue of our being in a situation (we are never not in one) and that in another situation the discrimination will also have already been made, but differently. In other words, while at any one point it is always possible to order and rank "Is there a text in this class?"1 and "Is there a text in this class?"2 (because they will always have already been ranked), it will never be possible to give them an immutable once-and-for-all ranking, a ranking that is independent of their appearance or nonappearance in situations (because it is only in situations that they do or do riot appear).
Nevertheless, there is a distinction to he made between the two that allows us to say that, in a limited sense, one is more normal than the other: for while each is perfectly normal in the context in which their literalness is immediately obvious (the successive contexts occupied by my colleague), as things stand now, one of those contexts is surely more available, and therefore more likely to be the perspective within which the utterance is heard, than the other. Indeed, we seem to have here an instance of what I would call "institutional nesting": If "Is there a text in this class?", is hearable only by those who know what is included under the rubric "first day of class," and if "Is there a text in this class?", is hearable only by those whose categories of understanding include the concerns of contemporary literary theory, then it is obvious that in a random population presented with the utterance, more people would "hear" "is there a text in this class?"1, than "Is there a text in this class?"2; and, moreover, that while "Is there a text in this class?"1 could be immediately hearable by someone for whom "IS there a text in this class?"2 would have to be laboriously explained, it is difficult to imagine someone capable of hearing "Is there a text in this class?"2 who was not already capable of hearing "Is there a text in this class."1 (One is hearable by anyone in the profession and by most students and by many workers in the book trade, and the other in the profession who would not think it peculiar to find, as I did recently, a critic referring to a phrase "made popular by Lacan."). To admit as much is not to weaken my argument by reinstating the category of the normal, because the category as it appears in that argument is not transcendental but institutional; and while no institution is so universally in force and so perdurable that the meanings it enables will be normal for ever, some institutions or forms of life are so widely lived in that for a great many people the meaning they enable seem "naturally" available and it takes a special effort to see that they are products of circumstances.
The point is an important one, because it accounts for the success with which an Abrams or an E. D. Hirsch can appeal to a shared understanding of ordinary language and argue from that understanding to the availability of a core of determinate meanings. When Hirsch offers "The air is crisp" as an example of a "verbal meaning" that is accessible to all speakers of the language, and distinguishes what is sharable and determinate about it from the associations that may, in certain circumstances, accompany it (for example, "I should have eaten less at supper," "Crisp air reminds me of my childhood in Vermont"), he is counting on his readers to agree so completely with his sense of what that shared and normative verbal meaning is that he does not bother even to specify it; and although I have not taken a survey, I would venture to guess that his optimism, with respect to this particular example, is well founded. That is, most, if not all, of his readers immediately understand the utterance as a rough meteorological description predicting a certain quality of the local atmosphere. But the "happiness" of the example, far from making Hirsch's point (which is always, as he has recently reaffirmed, to maintain "the stable determinacy of meaning") makes mine. The obviousness of the utterance's meaning is not a function of the values its words have in a linguistic system that is independent of context; rather, it is because the Words are heard as already embedded in a context that they have a meaning that Hirsch can then cite as obvious. One can see this by embedding the words in another context and observing how quickly another "obvious" meaning emerges. Suppose, for example, we came upon "The air is crisp" (which you are even now hearing as Hirsch assumes you hear it) in the middle of a discussion of music ("When the pieces played correctly the air is crisp"): it would immediately be heard as a comment on the performance by an instrument or instruments of a musical air. Moreover, it would only be heard that way, and to hear it in Hirsch's way would require all effort on the order of a strain. It could be objected that in Hirsch's text "The air is crisp", has no contextual setting at all; it is merely presented, and therefore any agreement as to its meaning must be because of the utterance's acontextual properties. But there is a contextual setting and the sign of its presence is precisely the absence of any reference to it. That is, it is impossible even to think of a sentence independently of a context, and when we are asked to consider a sentence for which no context has been specified, we will automatically hear it the context in which it has been most often encountered. Thus Hirsch invokes a context by not invoking it: by not surrounding the utterance with circumstances, he directs us to imagine it in the circumstances in which it is most likely to have been produced: and to so imagine it is already to have given it a shape that seems at the moment to be the only one possible
What conclusions can be drawn from these two examples? First of all, neither my colleague nor the reader of Hirsch's sentence is constrained by the meanings words have in a normative linguistic system; and yet neither is free to confer on an utterance any meaning he likes. Indeed, "confer" is exactly the wrong word because it implies a two-stage procedure in which a reader or hearer first scrutinizes an utterance and then gives it a meaning. The argument of the preceding pages can be reduced to the assertion that there is no such first stage, that one hears an utterance within, and not as preliminary to determining, a knowledge of its purposes and concerns, and that to so hear it is already to have assigned it a shape and given it a meaning. In other words, the problem of how meaning is determined is only a problem if there is a point at which its determination has not yet been made, and I am saying that there is not such point.